Perianesthetic hypothermia is common and proven to cause numerous serious complications. The investigators propose to evaluate two clinical observations: 1) the pattern of intraoperative hypothermia differs substantially in infants, young adults, and the elderly; and, 2) patients undergoing epidural or spinal anesthesia become 1-3 degrees C hypothermic before triggering shivering. Preliminary data suggest that unexpected physiological mechanisms underlie these observations. After several hours of surgery, core temperature in young adults stops decreasing and subsequently remains constant. An increase in metabolic heat production does not contribute to this core temperature plateau and, surprisingly, decreased cutaneous heat loss also contributes little. Accordingly, the investigators propose to test the hypothesis that the core temperature plateau in young adults results from constraint of metabolic heat to the core compartment by thermoregulatory vasoconstriction. Unwarmed infants initially become hypothermic during surgery and when sufficiently hypothermic, trigger thermoregulatory vasoconstriction. However, they subsequently increase their core temperatures, sometimes returning to control values. The investigators propose to test the hypothesis that nonshivering thermogenesis contributes significantly to the increase. In contrast, the elderly continue to become hypothermic during surgery and consequently have lower postoperative core temperatures. The investigators propose to test the hypothesis that the elderly continue to become hypothermic because they do not trigger thermoregulatory vasoconstriction during anesthesia. Patients undergoing epidural or spinal anesthesia become 1-3 degrees C hypothermic before triggering shivering (above the block). Since the brain is not directly influenced by regional anesthesia, it remains unclear why protective thermoregulatory responses should be impaired during conduction anesthesia. Preliminary data indicate that a leg skin temperature near 40 degrees C is required in unanesthetized volunteers to produce cold tolerance comparable to that produced by epidural anesthesia. The investigators, therefore, propose to test the hypothesis that thermoregulatory responses to core hypothermia are impaired because regional anesthesia increases apparent lower body temperature far more than it increases actual tissue temperature. These studies will identify the age-dependent efficacy of intraoperative thermoregulatory responses. Preliminary data suggest that responses are remarkably efficient in infants and virtually ineffective in the elderly. Confirming this hypothesis would indicate that the elderly require especially aggressive temperature monitoring and thermal management during general anesthesia. The investigators also will determine the extent to which thermoregulatory control is impaired during epidural anesthesia, and evaluate its etiology. Combined, these studies will facilitate optimal thermal management of surgical patients.